Commentary on 'Preliminary Species Hypotheses' in Entomological Taxonomy: A Global Data and FAIR Infrastructure Perspective

Biodiversity Data Journal 13: e141562
doi: 10.3897/BDJ.13.e141562

Commentary on "Preliminary Species Hypotheses"
in Entomological Taxonomy: A Global Data and
FAIR Infrastructure Perspective

Sharif Islam * §

+ Naturalis Biodiversity Center, Leiden, Netherlands
§ DiSSCo, Leiden, Netherlands

Corresponding author: Sharif Islam (sharif.islam@naturalis .nl)

Academic editor: Lyubomir Penev

Received: 11 Nov 2024 | Accepted: 21 Jan 2025 | Published: 10 Feb 2025

Citation: Islam S (2025) Commentary on "Preliminary Species Hypotheses" in Entomological Taxonomy: A
Global Data and FAIR Infrastructure Perspective. Biodiversity Data Journal 13: e141562.

https ://doi.org/10.3897/BDJ.13.e141562

Abstract

What if early taxonomic findings were treated like preprints, open to iterative improvement
or managed with practices from the open-source community, such as Git branching,
merging and patch management? Prompted by Buckley's article Charting a Future for
Entomological Taxonomy in New Zealand (2024), this commentary explores these
possibilities in the context of biodiversity informatics. In response to the need for rapid,
scalable biodiversity monitoring, Buckley introduces preliminary species hypotheses
(PSH) as a bridge between quick identification tools and the rigorous Linnaean system,
leveraging DNA barcoding and Al-assisted image recognition to produce provisional
Classifications that can later be validated. Expanding on Buckley’s framework, this
commentary emphasises the critical role of data linking, versioning and integration to
support evolving taxonomic data. Borrowing from software and open-source practices, |
explore the idea of managing PSH with an infrastructure that treats each taxonomic
update as a versioned "commit", which can be tracked, refined and integrated over time.
Drawing insights from FAIR (Findable, Accessible, Interoperable, Reusable) principles
and Digital Extended Specimens, | identify infrastructure requirements for PSH, including
robust data standards, persistent identifiers and interoperability to support global
biodiversity repositories. Additionally, Taxonomic Data Objects offer a model for

©@lslam S. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are
credited.

2 Islam S

dynamically integrating PSH into adaptable taxonomies that can evolve with new data
and tools. By positioning PSH within an open, infrastructure-focused framework, this
commentary advocates for scalable, hypothesis-driven biodiversity data that meets
modern conservation needs, bridging traditional and emerging practices in taxonomy.

Keywords

taxonomy, species, interoperability, FAIR, data integration, open source

Introduction

In Charting a Future for Entomological Taxonomy in New Zealand, published in the
journal New Zealand Entomologist, T.R. Buckley (2024) proposes the concept of
preliminary species hypotheses (PSH) as a way to bridge the gap between the need for
rapid species identification and the rigorous Linnaean taxonomy. Buckley argues that
PSH can address biodiversity monitoring needs by utilising output of rapid identification
tools -- such as DNA barcoding and Al-assisted image recognition - as provisional
Classifications that serve as an intermediate stage before formal taxonomic classification.
Although based in New Zealand and focused on entomology, this proposal has
implications for other regions and fields within taxonomy and biodiversity research.

Buckley's proposal envisions scalable, hypothesis-driven biodiversity data that can
evolve as new information emerges. Inspired by this approach, we might ask: What if
early taxonomic findings were treated like preprints (Verma and Detsky 2020) - open to
iterative improvement? Or managed through practices adapted from open-source
software development, such as Git branching, merging and patch management, where
each PSH acts as a versioned "commit"? While this iterative process echoes how
taxonomic science has traditionally progressed, this approach could offer a flexible
framework for tracking and refining taxonomic data over time.

In this commentary, | explore the data linking and integration infrastructure required to
support Buckley’s vision, emphasising how PSH fits within the broader framework of the
Species Hypotheses (SH) and Taxon Hypotheses (TH) paradigm. These concepts also
align with evolving Biodiversity Information Standards (TDWG) standards like the Taxon
Concept Schema (TCS), which separates Taxon Concepts from Taxon Names to
enhance data interoperability (Klazenga and Liljeblad 2024). Based on these conceptual
frameworks, | explore the data linking and integration infrastructure required to support
Buckley's vision, drawing on knowledge infrastructure studies such as Christina
Borgman's work on data systems (Borgman and Brand 2024) and Sterner et al.'s
pluralistic framework for biodiversity data sharing (Sterner et al. 2023). | also consider
recent proposals, such as Digital Extended Specimens (Hardisty et al. 2022) and
Taxonomic Data Objects (Upham and Poelen 2024), as potential models for integrating
PSH and other hypotheses-driven insights (both from molecular- and media-based
multimodal workflow) as data products within global biodiversity infrastructures. This

infrastructure-based approach can help sustain taxonomy's relevance in conservation
and research. My focus is not on assessing the scientific rigour of PSH, but rather on the
data linking and integration strategies that could underpin its implementation, offering a
scalable pathway for the evolution of taxonomic knowledge.

Summary of paper

Buckley’s proposal introduces PSH as a practical and flexible way to address the gap
between rapid biodiversity identification needs and the more formal Linnean
classification system. The paper presents the proposal with a historical background of
entomological taxonomy in New Zealand, discussing the reasons for declining taxonomy
funding and the importance of maintaining scientific rigour. For relevance, the summary
here highlights the key concepts of PSH.

This provisional approach of PSH aligns well with similar concepts, such as Operational
Taxonomic Units (OTUs) in DNA barcoding, which serve as proxies to categorise
unidentified taxa for integration across different biodiversity datasets and use cases.
According to Buckley, OTUs are typically molecular-based groupings (often not derived
from DNA-sequenced specimens, particularly for environmental DNA data). As Buckley
(2024):9 states:

“_.it is difficult to reconcile these OTUs [OTUs that are not derived from DNA
sequencing specimens and do not have a physical reference specimen] with
other types of character data. From a hypothesis testing perspective, these OTUs
can also be considered ‘preliminary species hypotheses’, but with a weaker
degree of support than from specimen-based DNA sequencing approaches (as
outlined earlier). This approach will require a large-scale eDNA survey of New
Zealand, focusing on the sampling of soil, water, air and insect trap residues.
Achieving this goal would also be a 5- to 10-year project with a moderate financial
investment. The output would be a comprehensive database of OTUs that, over
time, could be connected to described species or to DNA sequences obtained
from individual specimens".

In contrast, PSH are structured as an intermediate classification that is less formal than
Linnaean taxonomy, but aspires to achieve it over time. Unlike OTUs, PSH are not simply
molecular clusters; they are hypotheses that can later be validated and incorporated into
formal taxonomy as additional data become available. Buckley also reminds us in the
paper that similar methods are commonly used in fields of mycology (KOljalg et al. 2013)
and bacteriology. While OTUs offer a rapid and flexible tool for biodiversity estimation,
PSH are designed to be a step closer to formal species recognition, enabling hypothesis-
driven research and prioritisation without bypassing rigorous taxonomic standards
entirely:

“The goal is not to replace the Linnean system, or to lower its scientific
robustness, but to provide a framework for describing biodiversity more quickly

4 Islam S

than Linnean taxonomy can. DNA data can characterise lineages that, in turn, can
be considered as ‘preliminary species hypotheses’. These hypothesised species
can be tested, verified and described by taxonomists later if resources become
available. In the meantime, these hypothesised species can be used as a basis in
downstream conservation actions or ecological studies that require biodiversity to
be divided into scientifically meaningful entities. However, it must be remembered
that these hypothesised species have not been subject to robust testing and,
therefore, any downstream inference will not be as reliable as that from a fully
revised taxon” (Buckley 2024:8).

Furthermore, the robustness and benefit of the hypothesis-driven and iterative approach
come not just from a single data type, but from integrating a variety of data types. For
instance, combining molecular-based methods and multi-modal Al techniques can
significantly reduce uncertainties in the inference of observations:

"If we want robust species hypotheses, then large numbers of characters will
continue to be needed. There are technologies emerging that promise to greatly
increase the rate of data collection without sacrificing scientific robustness. The
approach adopting these technologies is known as /arge-scale integrative
taxonomy (Hartop et al. 2022; Salili-James et al. 2023; Karbstein et al. 2024).
Briefly, this approach comprises two steps. First, high throughput methods are
used to collect character data and perform a provisional grouping of specimens
into putative species. Second, another character type, with a high a pron
probability of being incongruent with the first character set, is used to test those
putative species (Hartop et al. 2022). A key feature is the use of technology to
accelerate the rate and scale of data collection" (Buckley 2024:7).

The demand for taxonomic information for a variety of use cases (Such as environmental
monitoring and biosecurity) is rising, making traditional insect sampling and identification
methods increasingly impractical, especially amidst a shortage of experts. New
technologies, including DNA barcoding, eDNA for community assessments and
automated image recognition, offer promising alternatives that can democratise species
identification. Automated image recognition, in particular, enables non-specialists to
identify insects, making taxonomy more accessible. However, according to Buckley,
successful adoption of these tools requires extensive digitisation of specimen records
and integration with images, DNA sequences and geo-referenced data.

Key Terms, Definitions and Alignment with Existing Concepts

The practice of taxonomy and nomenclature deals with different concepts and terms
beyond naming species (see Favret (2024) for 5 'D's of taxonomy: delimitation, diagnosis,
description, determintation and discovery) where the aspect of testable hypotheses
intersects all of these concepts. While detailing every aspect is beyond the scope of this
commentary, this section defines key terms and situates them within the evolving
landscape of biodiversity informatics. The following concepts are briefly stated here to

facilitate the discussion and lay the foundation for understanding how PSH can integrate
into taxonomic workflows and biodiversity data infrastructures.

Barcode Index Numbers (BINs): BINs are molecular-based clusters derived from DNA
barcoding, primarily serving as proxies for species identification using genetic
divergence thresholds. BINs are similar to OTUs, but are specific to DNA barcoding.
Unlike OTUs, which are often used as an intermediate step requiring further species-level
identification, BINS are dynamic and the boundaries of what sequences can be
associated with a particular BIN can change with new sampling data (Ratnasingham and
Hebert 2013; Lue et al. 2022) and one BIN can cover more than one taxon (Huemer and
Mutanen 2022).

Species Hypotheses (SH): SH is the main building block of UNITE (a database and
sequence management environment centred on the eukaryotic nuclear ribosomal ITS
region) which groups similar sequences into provisional species-level clusters typically
comprising two or more sequences to avoid excessive inflation (Kdljalg et al. 2013).
Representative sequences for each SH are chosen through consensus computation or
expert designation. These SHs, along with their representative sequences and
annotations, are made available as reference datasets. Buckley's paper discusses SH
used in mycology and explores how entomology can adapt similar ideas. This discussion
also opens up the possibility of integrating SH concepts for broader use beyond
mycology and zoology, not necessarily limited to DNA-based identification methods.

Taxon Hypotheses (TH) paradigm: Expanding on the SH concept, Koljalg et al. (2020)
introduces the TH paradigm that represents a framework for linking sequence-based
identifications to taxonomic concepts. By assigning Digital Object Identifiers (DOls) to
these hypotheses, THs enable transparent and reproducible connections between
molecular data and taxonomic classifications. Kdljalg et al. (2020) also highlights that,
while molecular data are becoming increasingly common, differences in sampling,
genetic markers and analytical methods often lead to competing and sometimes
conflicting classifications. The reference datasets and DOls provided by UNITE offer a
unique reference point that remains consistent even as underlying data and conclusions
evolve. This system allows users to reference the data enabling modifications and
augmentations, while preserving original versions.

All of these frameworks have one thing in common: they acknowledge the dynamic and
"preliminary" nature of initial insights into species identification. Thus, PSH or SH could
emerge as a new "data type" that can be used not just in mycology or zoology, but across
domains. Furthermore, this approach supports integrative methods that apply multiple
types of characters, leading to robust hypothesis tests and, therefore, greater confidence
in the acceptance or rejection of a species hypothesis.

Recent discussions (see Karbstein et al. (2024)) on species delimitation and Al also
underscore the importance of incorporating multiple data types and frameworks such as
unified species concept, morphological and phylogenetic (genetic relationships and
shared ancestry) and DNA clustering methods that are going towards a more integrative

6 Islam S

approach (genetics/genomics + morphology + ecology). Al-based identification methods,
including multimodal approaches involving sound and vision, are also becoming
increasingly prevalent (Waldchen and Mader 2018; Yang et al. 2021). Each approach
has limitations; thus, integrative approaches that combine multiple lines of evidence align
with the dynamic nature of species hypotheses.

By situating PSH, SH, TH, BINs and OTUs within a unified conceptual framework, this
commentary underscores the value of treating species hypotheses as dynamic, evolving
data objects. Each concept - BINs, OTUs, SH, TH and PSH - has distinct origins rooted in
specific fields, such as molecular biology, fungal taxonomy and entomology. These
approaches complement the Linnaean classification by integrating preliminary taxonomic
data into an iterative process that refines and validates hypotheses over time. Expanding
their application to encompass diverse data types will enhance their utility across
taxonomic domains. A holistic and integrative approach supports the iterative refinement
of taxonomies while balancing the need for rapid discovery with the production of robust,
high-quality data.

The role of infrastructures

Following the summary of Buckley’s PSH proposal, it becomes clear that data integration
and linking will be an important aspect and, thus, the successful implementation and
sustainability of PSH require a robust digital infrastructure. This infrastructure not only
enables data sharing, but also supports the evolution of taxonomic knowledge in a
scalable and accessible way. The PSH model is comparable to preprints in scholarly
publishing: it provides a way to make new insights accessible, citable and linkable, even
if they require further refinement and validation. When viewed through the lens of the
Digital Extended Specimen (DES) paradigm (Hardisty et al. 2022) and the FAIR
(Findable, Accessible, Interoperable, Reusable) principles, the PSH concept highlights
the need for infrastructure that can support both provisional classifications and long-term
taxonomic research. The intersection of PSH with DES and FAIR principles underscores
the challenges - and critical importance - of establishing, maintaining and scaling digital
infrastructure to meet the demands of modern biodiversity research. This is not to argue
for a new type of digital infrastructure, but improving on existing infrastructures and
aligning global and regional funding schemes that can be adopted to implement such a
proposal. Similar to Buckley, Meier et al. (2024) also emphasise that achieving
integrative taxonomy (combining morphological, whole organism study with molecular
data) requires reliable data handling, including efficient voucher storage, standardised
data practices and FAIR-compliant infrastructure to support the evolution of taxonomic
hypotheses as new data are added.

For biodiversity data to be effective, including taxonomic and nomenclature information, a
resilient infrastructure is crucial to maintain links amongst evolving species hypotheses,
underlying specimens, environmental observations and genetic data. Efforts to create
such infrastructures have accelerated globally as we confront biodiversity and climate
crises (Devictor and Bensaude-Vincent 2016). Although global data infrastructures that

support biodiversity data and research funding are unevenly distributed, the DES and
PSH approach could mitigate disparities by providing an inclusive, interoperable system
that enables biodiversity data sharing across regions and disciplines.

The DES, as proposed, is a paradigm for digitally linking specimen data from global
natural science collections to related taxonomic, ecological and environmental data. DES
enables the transformation of physical specimen data into digital objects, making them
accessible and FAIR. This approach not only broadens usability, but also enhances the
value of collections by integrating them into global data infrastructures that can be
leveraged for large-scale, multifactor analysis (Heberling et al. 2021). Thinking about
DES, PSH and FAIR in a holistic framework brings up the notion of pluralistic data
pooling advocated by Sterner et al. (2023):2:

We define ‘data pooling’ for biodiversity data as a process that combines data
from multiple sources into one taxonomically standardized body of information,
provides infrastructure for managing and accessing the combined data and
governs it as a shared resource for a community of users and stakeholders
beyond a single research project or lab. We define ‘taxonomic standardization’ as
a set of processes for verifying and re-identifying a collection of species
observations as needed to ensure that they are classified in a standardized way
according to a single, coherent taxonomy of choice. More generally, ‘data
standardization’ (also Known as data harmonization) is an established term in
academic and industry data science practices.

Part of this set of process can be a PSH data element that can accommodate evolving
taxonomic concepts, while ensuring reliable links between data sources. It allows for both
the robustness of Linnean taxonomy and the flexibility of documenting hypotheses,
thereby fostering a dynamic approach to biodiversity research. Echoing Sterner (also
Leonelli (2020) and Borgman and Wofford (2021)), the challenges of biodiversity data
collection, sharing and preservation are as much social as technical, thus:

“...making biodiversity data comprehensively available and reusable will likely
require major changes to the cultures, organizations and infrastructures of the
research communities involved” (Sterner et al. 2023: 2).

This also brings up the notion of maintenance and support. As Borgman et al. (2016)
note, "durability" in infrastructure requires continuous maintenance across technical and
human resources. Applying this insight to biodiversity data infrastructure highlights that
building a sustainable, FAIR-compliant system requires not only technical innovation, but
also governance and investment. Borgman’s work in astronomy shows that even well-
established systems still face fragility without regular support - an important reminder as
we build infrastructures that will support biodiversity data on a global scale.

8 Islam S

Integration with Global Data Standards and Networks

As mentioned already, PSH can expand beyond New Zealand and entomology; it has
potential for integration with global biodiversity data initiatives. Organisations and
platforms such as the Catalogue of Life, GBIF, BCON, ALA, INSDC, BOLD, UNITE and Di
SSCo provide frameworks, tools and services for aggregating and curating biodiversity
data, which could be expanded to incorporate PSH as a new type of digital object. By
embedding provisional species data into the global biodiversity network, PSH could
become widely accessible and actionable across regions and disciplines.

As Moersberger et al. (2024) emphasise in their study on European biodiversity
monitoring, integrating biodiversity data is crucial for reducing fragmentation and filling
taxonomic gaps. Aligning PSH with the shift toward digital taxonomy could further bridge
the divide between morphological and molecular approaches, providing traceable,
reusable links to each hypothesis’s provenance. This would enable a more cohesive and
adaptable taxonomy, supporting dynamic updates as new data become available.

Enhancing PSH with FAIR Compliance

To fully realise PSH, we need infrastructure that is both accessible and FAIR-compliant.
These hypotheses will function as data points or nodes within a knowledge graph (Page
2019,Penev et al. 2024) and, because they could be stored across multiple
infrastructures (Sterner et al. 2020), data linking and interoperability are essential. The
Upham and Poelen (2024) concept of Taxonomic Data Objects aligns with this need by
offering machine-readable digital packages that encode metadata, enabling the tracking
of evolving species concepts over time. Initial taxonomic data can also be compared to a
software commit in Git: each PSH represents a specific "state" of species classification,
preserving the evolution of taxonomic understanding without overwriting earlier
hypotheses. This approach provides a clear pathway for reviewing and merging
provisional classifications with established taxonomies, strengthening taxonomic
workflows by ensuring data integrity and interoperability across different taxonomic
systems (see Fig. 1 for a simple schematic comparing Git merging with the process
described using PSH).

Practical Requirements for Preliminary Species Hypotheses
Implementation

For PSH to serve as a valuable tool in taxonomy and biodiversity informatics, certain key
elements are essential. This is an initial proposal and will benefit from further discussion:

1. Persistent Identifiers (PIDs): Each PSH digital object should be assigned a PID
to ensure reliable tracking and referencing, similar to the approach used for
Digital Extended Specimens within the FAIR Digital Object framework (Islam et al.
2023). As suggested by Upham and Poelen (Upham and Poelen 2024),

versioning and hashing could be incorporated as part of the metadata to support
tracking changes over time. Assigning PIDs to taxonomic data and hypotheses is
not a new concept; for example, the Catalogue of Life assigns identifiers for name
usage and checklists (Banki et al. 2023) and UNITE assigns DOls to species
hypotheses (KOljalg et al. 2020). The discussion should not focus on which
specific PID mechanism is optimal - though implementation details are important -
but rather on establishing a consensus and actionable plan to assign PIDs to
these entities at a granular level. This will enable effective tracking and linking,
but requiring dedicated infrastructure and ongoing maintenance support. By
assigning transparent and persistent identifiers to contributors across all stages of
a species hypothesis’ evolution, the framework could foster equitable recognition
while maintaining rigorous standards for formal naming.

Taxonomic Workflow

Specimen Collection & Environmental DNA & Multi

Git Ideas Morphological Data Modal Data
Main Taxonomic Data | |
Branch Initial Identification initial Identification
? i |
/ \
{ \ Species Hypothesis: SH2 Species Hypothesis: SH1
’ ,
Hypotheses Branch 2 Hypotheses Branch 1 %
= ye
/ ~ ~
4 SH Integrative Data
a Pe
Hypotheses Sub-branch 2.1
¥
Provisional Data Integration
¥
|
Combined Hypotheses
Branch |
¥
FAIR and Data Standards
Compliance/ Taxon
’ Concept Schema Usage
FAIR and Data Standards |
Alignment/ Taxon Concept |
Schema Usage Linked with PIDs
¥
Linked with PIDs Long term Data Repository
’
Long Term Data Repository |
®
Revised SH with Additional
Reviewed & Accepted Data

v |
Refined Taxonomic Data Reviewed & Accepted
¥

Refined Taxonomic Data

Figure 1. EE

A simplified conceptual framework for version-controlled taxonomic data
management

This diagram illustrates the parallel between hypothesis-driven taxonomic workflows and Git-
based version control systems. Drawing inspiration from software development practices, the
framework demonstrates how version control concepts could be applied to manage and track
the evolution of taxonomic hypotheses. The actual processes involved are much more
complex, as described in Pyle's paper "An Introduction to Scientific Names of Organisms and
the Taxon Concepts they Represent (Pyle 2022).

10

Islam S

Interoperable Data Standards: Standards like Darwin Core and Taxon Concept
Schema (TCS) are necessary to harmonise species hypothesis data with other
biodiversity data types, such as observation and occurrence data. Consistent
standards enable smoother integration and reuse of taxonomic information across
platforms. How a preliminary concept could be part of Darwin Core and other
standards framework will need careful consideration. For instance, “dwc:previous
Identifications” property in Darwin Core could store the reference to preliminary
data . PSH, SH and TH could have their own data model and metadata, but this
also needs global consensus. As new data and insights are being generated,
standards and schemas are essential for usability in diverse contexts. While
Darwin Core is widely used, TCS’s separation of Taxon Concepts from Taxon
Names allows greater flexibility for mapping and resolving taxonomic data. TCS
could possibly accommodate dynamic states such as "Preliminary" and "Final" as
new insights emerge. It could also address provenance and attribution, akin to the
Linnaean tradition of authorship, requiring each state to have a source
("according To") (Klazenga and Liljeblad 2024).

FAIR Principles: Along with PIDs, machine-readable formats and data standards
will enhance accessibility, interoperability and reusability, supporting transparent
and evolving taxonomic classifications. Similar ideas have been proposed by
Miralles et al. (2020) in the context of alpha taxonomy repositories. Taxonomic
Data Objects (Upham and Poelen 2024) could standardise PSH data in a
machine-readable format, preserving their structure and allowing flexible data
use.

Global Coordination and open source practices: Collaborative efforts with
established networks are essential for integrating PSH into a global biodiversity
framework. Beyond achieving consensus on metadata standards, the accessibility
and publication of these data must remain a priority. Funders, research institutions
and collection-holding organisations need to recognise the importance of APIs (
Addink et al. 2023), repositories, data stewardship (De Prins 2019Bentley et al.
2024) and other foundational infrastructure and commit both human and
technological resources to support them. This is especially crucial given that
many countries, despite their reliance on biodiversity data for modelling and
monitoring, often lack the necessary capacity, expertise or funding to fully exploit
its potential (WMoersberger et al. 2024). As illustrated by New Zealand's example,
where a small population and limited taxonomic expertise hinder the
development of comprehensive taxonomic research, many countries depend on
international collaboration for taxonomic knowledge. Addressing this taxonomic
impediment calls for capacity building, knowledge exchange and the creation of
sustainable, FAIR-aligned taxonomic services through coordinated efforts (
Buckley 2024). A unified global solution may be impractical, yet stronger
coordination in the software and standards that support taxonomic services is
critical. This can facilitate the effective use of new data elements like PSH and
promote shared governance structures. For instance, the discussions by Sandall
et al. (2023) on checklist maintenance can be extended to taxonomic software
and service development, where PSH could be tested and refined. Capacity

11

management and funding challenges also require open dialogue, especially
given the voluntary nature of many contributions in taxonomy and also in
biodiversity informatics and data stewardship. Metrics from open-source projects,
such as the "Contributor Absence Factor" (or "Bus Factor") - which assesses how
many contributors can be lost before a project is impacted - could help guide
efforts towards sustainability. By learning from open-source practices and
research software sustainability principles (Cohen et al. 2021), we can enhance
taxonomy's resilience and interoperability across regions. While taxonomic
expertise remains indispensable, adopting insights from open-source and other
data ecosystems will help us to overcome challenges in data infrastructure and
interoperability.

Conclusion

Buckley's concept of PSH, primarily proposed within entomology, parallels existing
frameworks like SH in mycology and BINs and OTUs from molecular methods. Despite
their overlaps and distinctions, the need for standardised frameworks to manage
preliminary and evolving taxonomic data remains crucial. These frameworks address
challenges across diverse taxonomic domains, emphasising their potential to create
interoperable and dynamic taxonomic practices, but a wider and global discussion is
needed to find a holistic solution.

In the context of New Zealand, Buckley advocates for shifting entomological taxonomy
away from the primary focus on completing Linnaean classification. Instead, his proposal
highlights achievable objectives aligned with realistic funding and _ timelines,
incorporating DNA data and Al methods as preliminary steps towards formal
classification. This commentary connects Buckley's proposal to broader initiatives, such
as FAIR principles, Digital Extended Specimens, Taxon Concept Schema, Taxonomic
Data Objects and open-source software practices. By treating PSH as data points -
similar to versioned git "commits" or "preprints" - species identification and classifications
can be iteratively refined without losing historical data. This fosters a more adaptable and
integrative approach to taxonomy, bridging morphological and molecular data and Al-
based identification, while enhancing global biodiversity conservation efforts.

Conflicts of interest

The authors have declared that no competing interests exist.

References

° Addink W, Kyriakopoulou N, Penev L, Fichtmueller D, Norton B, Shorthouse D (2023)
Deliverable D1.3 Best practice manual for findability, re-use and accessibility of
infrastructures. ARPHA Preprints_httos://doi.org/10.3897/arphapreprints.e107169

12

Islam S

Banki O, Déring M, Jeppesen T (2023) Name IDs and Name Matching for Catalogue of
Life: Existing Services and Prospects. Biodiversity Information Science and Standards 7
https ://doi.org/10.3897/biss.7.111662

Bentley A, Thiers B, Moser WE, Watkins-Colwell GJ, Zimkus BM, Monfils AK, Franz NM,
Bates JM, Boundy-Mills K, Lomas MW, Ellwood ER, Poo S, Contreras DL, Webster MS,
Nelson G, Pandey JL (2024) Community Action: Planning for Specimen Management in
Funding Proposals. BioScience 74 (7): 435-439. https://doi.org/10.1093/biosci/biae032
Borgman C, Sands A, Darch P, Golshan M (2016) The durability and fragility of
knowledge infrastructures: Lessons learned from astronomy. Proceedings of the
Association for Information Science and Technology 53 (1): 1-10. https://doi.org/10.1002/
pra2.2016.14505301057

Borgman C, Wofford M (2021) From Data Processes to Data Products: Knowledge
Infrastructures in Astronomy. Harvard Data Science Review https ://doi.org/
10.1162/99608f92.4e792052

Borgman C, Brand A (2024) The Future of Data in Research Publishing: From Nice to
Have to Need to Have? Harvard Data Science Review https://doi.org/
10.1162/99608f92.b73aae77

Buckley TR (2024) Charting a future for entomological taxonomy in New Zealand. New
Zealand Entomologist1-17. https://doi.org/10.1080/00779962.2024.2407230

Cohen J, Katz D, Barker M, Chue Hong N, Haines R, Jay C (2021) The Four Pillars of
Research Software Engineering. IEEE Software 38 (1): 97-105. httos://doi.org/10.1109/
ms .2020.2973362

De Prins J (2019) Global Open Biodiversity Data: Future Vision of FAIR Biodiversity
Data Access, Management, Use and Stewardship. Biodiversity Information Science and
Standards 3 https://doi.org/10.3897/biss.3.37190

Devictor V, Bensaude-Vincent B (2016) From ecological records to big data: the invention
of global biodiversity. History and Philosophy of the Life Sciences 38 (4). https://doi.org/
10.1007/s40656-016-0113-2

Favret C (2024) The 5 ‘D’s of Taxonomy: A User’s Guide. The Quarterly Review of
Biology 99 (3): 131-156. https ://doi.org/10.1086/732044

Hardisty AR, Ellwood ER, Nelson G, Zimkus B, Buschbom J, Addink W, Rabeler RK,
Bates J, Bentley A, Fortes JAB, Hansen S, Macklin JA, Mast AR, Miller JT, Monfils AK,
Paul DL, Wallis E, Webster M (2022) Digital Extended Specimens: Enabling an
Extensible Network of Biodiversity Data Records as Integrated Digital Objects on the
Internet. BioScience 72 (10): 978-987. https://doi.org/10.1093/biosci/biacO60

Heberling JM, Miller J, Noesgaard D, Weingart S, Schigel D (2021) Data integration
enables global biodiversity synthesis. Proceedings of the National Academy of Sciences
118 (6). https ://doi.org/10.1073/pnas.2018093118

Huemer P, Mutanen M (2022) An Incomplete European Barcode Library Has a Strong
Impact on the Identification Success of Lepidoptera from Greece. Diversity 14 (2). https://
doi.org/10.3390/d14020118

Islam S, Beach J, Ellwood E, Fortes J, Lannom L, Nelson G, Plale B (2023) Assessing
the FAIR Digital Object Framework for Global Biodiversity Research. Research Ideas
and Outcomes 9 https://doi.org/10.3897/rio.9.e108808

Karbstein K, Késters L, Hoda¢ L, Hofmann M, Hérandl E, Tomasello S, Wagner N,
Emerson B, Albach D, Scheu S, Bradler S, de Vries J, Irisarri 1, Li H, Soltis P, Mader P,
Waldchen J (2024) Species delimitation 4.0: integrative taxonomy meets artificial

13

intelligence. Trends in Ecology & Evolution 39 (8): 771-784. httos://doi.org/10.1016/j.tree.
2023.11.002

Klazenga N, Liljeblad J (2024) Expressing Circumscription in the Taxon Concept Schema
(TCS). Biodiversity Information Science and Standards 8 htips://doi.org/10.3897/biss.
8.140738

KOljalg U, Nilsson RH, Abarenkov K, Tedersoo L, Taylor AS, Bahram M, Bates S, Bruns
T, Bengtsson-Palme J, Callaghan T, Douglas B, Drenkhan T, Eberhardt U, Duenas M,
Grebenc T, Griffith G, Hartmann M, Kirk P, Kohout P, Larsson E, Lindahl B, LUcking R,
Martin M, Matheny PB, Nguyen N, Niskanen T, Oja J, Peay K, Peintner U, Peterson M,
Példmaa K, Saag L, Saar I, Schuler A, Scott J, Senés C, Smith M, Suija A, Taylor DL,
Telleria MT, Weiss M, Larsson K (2013) Towards a unified paradigm for sequence-based
identification of fungi. Molecular Ecology 22 (21): 5271-5277. htips://doi.org/10.1111/mec.
12481

K6ljalg U, Nilsson H, Schigel D, Tedersoo L, Larsson K, May T, Taylor AS, Jeppesen TS,
Frgslev TG, Lindahl B, Poldmaa K, Saar |, Suija A, Savchenko A, Yatsiuk |, Adojaan K,
Ivanov F, Piirmann T, P6hdnen R, Zirk A, Abarenkov K (2020) The Taxon Hypothesis
Paradigm—On the Unambiguous Detection and Communication of Taxa. Microorganisms
8 (12). https://doi.org/10.3390/microorganisms8121910

Leonelli S (2020) Learning from Data Journeys. Data Journeys in the Sciences1-24.

https ://doi.org/10.1007/978-3-030-37177-7_1

Lue C, Abram P, Hrcek J, Buffington M, Staniczenko PA (2022) Metabarcoding and
applied ecology with hyperdiverse organisms: Recommendations for biological control
research. Molecular Ecology 32 (23): 6461-6473. https://doi.org/10.1111/mec.16677
Meier R, Lawniczak MN, Srivathsan A (2024) Illuminating Entomological Dark Matter with
DNA Barcodes in an Era of Insect Decline, Deep Learning, and Genomics. Annual
Review of Entomology _https://doi.org/10.1146/annurev-ento-040124-014001

Miralles A, Bruy T, Wolcott K, Scherz MD, Begerow D, Beszteri B, Bonkowski M, Felden
J, Gemeinholzer B, Glaw F, Gléckner FO, Hawlitschek O, Kostadinov |, Nattkemper TW,
Printzen C, Renz J, Rybalka N, Stadler M, Weibulat T, Wilke T, Renner SS, Vences M
(2020) Repositories for Taxonomic Data: Where We Are and What is Missing. Systematic
Biology 69 (6): 1231-1253. https://doi.org/10.1093/sysbio/syaa026

Moersberger H, Valdez J, Martin JC, Junker J, Georgieva |, Bauer S, Beja P, Breeze T,
Fernandez M, Fernandez N, Brotons L, Jandt U, Bruelheide H, Kissling WD, Langer C,
Liquete C, Lumbierres M, Solheim AL, Maes J, Moran-Ordofiez A, Moreira F, Pe'er G,
Santana J, Shamoun-Baranes J, Smets B, Capinha C, McCallum |, Pereira H, Bonn A
(2024) Biodiversity monitoring in Europe: User and policy needs. Conservation Letters 17
(5). https://doi.org/10.1111/conl.13038

Page RM (2019) Ozymandias: a biodiversity knowledge graph. PeerJ 7 https://doi.org/
10.7717/peerj.6739

Penev L, Koureas D, Groom Q, Lanfear J, Agosti D, Casino A, Miller J, Cochrane G, Ba
n, O. K&, ljalg U, Ruch P (2024) Beyond BiCIKL: Towards Building an Al-Assisted
"Biodiversity Supergraph". Biodiversity Information Science and Standards 8: 135550.
https ://doi.org/10.3897/biss.8.135550

Pyle R (2022) An Introduction to Scientific Names of Organisms, and the Taxon Concepts
they Represent. Biodiversity Information Science and Standards 6 https://doi.org/10.3897/
biss.6.93926

14

Islam S

Ratnasingham S, Hebert PN (2013) A DNA-Based Registry for All Animal Species: The
Barcode Index Number (BIN) System. PLoS ONE 8 (7). https://doi.org/10.1371/
journal.pone.0066213

Sandall E, Maureaud A, Guralnick R, McGeoch M, Sica Y, Rogan M, Booher D, Edwards
R, Franz N, Ingenloff K, Lucas M, Marsh C, McGowan J, Pinkert S, Ranipeta A, Uetz P,
Wieczorek J, Jetz W (2023) A globally integrated structure of taxonomy to support
biodiversity science and conservation. Trends in Ecology & Evolution 38 (12):
1143-1153. https ://doi.org/10.1016/j.tree.2023.08.004

Sterner B, Gilbert E, Franz N (2020) Decentralized but Globally Coordinated Biodiversity
Data. Frontiers in Big Data 3 https ://doi.org/10.3389/fdata.2020.519133

Sterner B, Elliott S, Gilbert EE, Franz NM (2023) Unified and pluralistic ideals for data
sharing and reuse in biodiversity. Database 2023 https ://doi.org/10.1093/database/
baad048

Upham N, Poelen J (2024) Taxonomic Data Objects for Communicating the Meaning of
Species Names. Biodiversity Information Science and Standards 8 httos://doi.org/
10.3897/biss.8.139413

Verma A, Detsky A (2020) Preprints: a Timely Counterbalance for Big Data—Driven
Research. Journal of General Internal Medicine 35 (7): 2179-2181. httos://doi.org/10.1007/
$11606-020-05746-w

Waldchen J, Mader P (2018) Machine learning for image based species identification.
Methods in Ecology and Evolution 9 (11): 2216-2225. httos://doi.org/10.1111/2041-210x.
13075

Yang B, Zhang Z, Yang C, Wang Y, Orr MC, Wang H, Zhang A (2021) Identification of
Species by Combining Molecular and Morphological Data Using Convolutional Neural
Networks. Systematic Biology 71 (3): 690-705. https ://doi.org/10.1093/sysbio/syab076